Back

Comprehensive Analysis of RFID Performance of the iID BEEscience system

Magnay, J.; Cipriano, A. P.; Macro Versteeg, S.; Segers, F.; Grueter, C.

2025-06-06 animal behavior and cognition
10.1101/2025.06.03.657638 bioRxiv
Show abstract

Radio Frequency Identification (RFID) technology has been widely used to study the activity and behaviour of various organisms, including bees. In this study, we rigorously evaluate the performance of a recently developed RFID system (iID(R)BEEscience from MicroSensys) in tracking bees under different conditions. We assessed the systems accuracy in detecting and identifying travel direction both in controlled laboratory settings and with free-flying bees. Additionally, we analysed the sources of incorrect or redundant readings and provide a Python script designed to filter out erroneous data, summarising the results in biologically relevant measurements (foraging trip number, trip duration, instances of drifting). Our controlled laboratory tests revealed that the RFID systems accuracy ranged from 93% to 100% in most cases, though accuracy diminished when a large number of transponders passed through the reader simultaneously. We also found certain transponder positions in relation to the reader position to be less reliable. In field conditions, we tracked 33 foraging trips of RFID-tagged honeybees trained to a sucrose solution feeder and observed a 100% success rate in identifying all foraging trips via the RFID data. Given that the transponders weigh only 2.1 mg, we propose that this system is a reliable tool for studying smaller bees (and other insects), which are currently understudied using RFID technology, although its application to species smaller than honeybees (Apis mellifera) requires further testing. This research demonstrates the potential of RFID technology for advancing the study of insect behaviour under field conditions beyond traditional models like honeybees or bumblebees.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.